Carlos Guardiola

Industrial operators accumulate decades of telemetry, events and diagnostic records in legacy systems designed for supervision and certification rather than analytics. Recovering analytics-ready information from these brownfield archives is an industrial information integration problem: signals are fragmented across heterogeneous schemas, sparsely and asynchronously sampled, and detached from the operational context required for engineering interpretation. We formalise Industrial Data Archaeology (IDA) as the upstream problem class of non-disruptively reconstructing analytics-ready operational context from brownfield archives, and demarcate it from generic ETL, lakehouse ingestion, physical retrofit and clean-data predictive maintenance. IDA is specified as a typed archive instance, a reconstruction operator and eight architectural invariants that any solution must jointly satisfy, together with structural guarantees—boundary confinement of fleet-specific knowledge, idempotent windowed reconstruction and schema-evolution safety—under which onboarding a new fleet reduces to configuration. The methodology is instantiated as a non-disruptive Medallion retrofit and evaluated on a multi-fleet railway deployment spanning seven rolling-stock archetypes over 2.5 years. Across 376,012 orchestrated task instances (November 2025–June 2026) the pipeline sustains 99.99% final task success (41 final failures; 99.76% recovery on retry) with no memory-related terminations, absorbs 3,379 schema-evolution events without downtime, and achieves 8.44× compression on continuous telemetry (4.19 TB → 496 GB) together with a 1940.8× serving-cardinality reduction. Event-context enrichment reaches 70.75%, and every post-seed fleet onboarding was completed through configuration alone. The formalisation and invariants are domain-independent; consumption of the resulting Gold products by predictive-maintenance and digital-twin workloads is the natural downstream stage.

Nitrogen oxides (NOx) are regulated pollutants emitted by diesel engines, and accurate tailpipe NOx prediction is essential for engine calibration and regulatory compliance. Selective Catalytic Reduction (SCR) systems are widely used to reduce NOx emissions in modern diesel vehicles, yet reliable predictive modeling under real-world operating conditions remains challenging, particularly when balancing accuracy, interpretability, and model simplicity.This work is based on real-world measurements from a Euro 6 non-plug-in hybrid diesel vehicle equipped with an SCR system, considering nominal operating conditions. A control-oriented physico-chemical SCR model comprising 11 equations and 14 calibration parameters is calibrated using an asynchronous random Particle Swarm Optimization (arPSO) algorithm. Parameter identifiability is investigated through a numerical analysis of the distribution of low-error solutions. To improve predictive accuracy without increasing model order, a Hybrid Digital Twin (HDT) approach is proposed, in which a data-driven correction term models the residual between the physics-based predictions and measured NOx emissions. The correction is obtained using Grammatical Evolution with lexicase selection, resulting in an explicit analytical expression dependent on measured variables.While the calibrated physico-chemical model captures the main SCR dynamics, it exhibits systematic prediction errors. The proposed HDT significantly reduces these mismatches and improves NOx prediction accuracy on both training and independent test datasets, without compromising model transparency or computational efficiency.Hybrid Digital Twin framework effectively combines physical interpretability with enhanced predictive performance for tailpipe NOx emissions under nominal SCR operation. The approach provides a transparent and control-oriented solution suitable for benchmark simulations and model-based assessments of NOx compliance.

The Next-Generation IoT integrates diverse technological enablers, allowing the creation of advanced systems with increasingly complex requirements and maximizing the use of available IoT–edge–cloud resources. This paper introduces an orchestrator architecture for dynamic IoT scenarios, inspired by ETSI NFV MANO and Cloud Native principles, where distributed computing nodes often have unfixed and changing networking configurations. Unlike traditional approaches, this architecture also focuses on managing services across massively distributed mobile nodes, as demonstrated in the automotive use case presented. Apart from working as MANO framework, the proposed solution efficiently handles service lifecycle management in large fleets of vehicles without relying on public or static IP addresses for connectivity. Its modular, microservices-based approach ensures adaptability to emerging trends like Edge Native, WebAssembly and RISC-V, positioning it as a forward-looking innovation for IoT ecosystems.

Internal combustion engines must be individually tested at the end of the manufacturing process. In recent years classical hot test stands, where the engine is run for several minutes, are being replaced by cold test alternatives. The latter allow fast testing cycles using an external motoring device without using any fuel. The absence of fuel and combustion lowers the health and safety requirements for the plant itself and subsequent engine transport, but this comes at the cost of additional difficulties for the verification of the correct assembly and operation of the combustion system hardware.This paper presents a cold test concept, which includes dedicated measurements and algorithms for the detection of common failures in the manufacturing process, including those of the combustion hardware. Between other parameters, the test stand evaluates the existence of compression leaks, the correct variable valve timing (VVT) phasing, the operation of the ignition coils and injectors, and verifies the spark plugs gap to be within specifications.In all cases, non-intrusive sensors are used and the installation is done taking advantage of the standard wiring of the engine. Compression leaks are detected using the motoring torque and engine speed, and performing a frequency domain analysis of the signal; VVT phasing is verified by camshaft and crankshaft encoders, and with the recognition of the pulsation of the pressure in the manifolds; for the case of ignition coils and injectors, the system is electrically actuated and pattern recognition is applied to the intensity profile; finally, spark plug gap is inferred from the spark success rate at different energizing times.